[Intel-gfx] [PATCH v6] drm/i915: Fix FBC cfb stride programming for non X-tiled FB

[Praveen Paneri]

When FBC is enabled for linear, legacy Y-tiled and Yf-tiled
surfaces on gen9, the cfb stride must be programmed by SW as

cfb_stride = ceiling[(at least plane width in pixels)/
		     (32 * compression limit factor)] * 8

v2: Minor fix for a build error
v3: Fixed subject, register name and platform check (Ville)
v4: Added WA details in comment (Paulo)
v5:
 - Read modified reg write to preserve other bit values (Paulo)
 - Store modified stride value in reg_params (Paulo)
 - Keep GLK out of the WA (Paulo)
v6:
 - added additional field in reg_params for gen9_wa_cfb_stride (Paulo)
 - Used appropriate bit mask while writing the register (Paulo)

Cc: Paulo Zanoni <paulo.r.zanoni at intel.com>
Signed-off-by: Praveen Paneri <praveen.paneri at intel.com>
---
 drivers/gpu/drm/i915/i915_drv.h  |  1 +
 drivers/gpu/drm/i915/i915_reg.h  |  4 ++++
 drivers/gpu/drm/i915/intel_fbc.c | 27 +++++++++++++++++++++++++++
 3 files changed, 32 insertions(+)

diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 907603c..1d40a7c 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1106,6 +1106,7 @@ struct intel_fbc {
 		} fb;
 
 		int cfb_size;
+		unsigned int gen9_wa_cfb_stride;
 	} params;
 
 	struct intel_fbc_work {
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 56df86e..51cab2f 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -6801,6 +6801,10 @@ enum {
 #define  GLK_CL1_PWR_DOWN	(1 << 11)
 #define  GLK_CL0_PWR_DOWN	(1 << 10)
 
+#define CHICKEN_MISC_4		_MMIO(0x4208c)
+#define   FBC_STRIDE_OVERRIDE	(1<<13)
+#define   FBC_STRIDE_MASK	0x1FFF
+
 #define _CHICKEN_PIPESL_1_A	0x420b0
 #define _CHICKEN_PIPESL_1_B	0x420b4
 #define  HSW_FBCQ_DIS			(1 << 22)
diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c
index 860b8c2..21f6b33 100644
--- a/drivers/gpu/drm/i915/intel_fbc.c
+++ b/drivers/gpu/drm/i915/intel_fbc.c
@@ -291,6 +291,21 @@ static void gen7_fbc_activate(struct drm_i915_private *dev_priv)
 	u32 dpfc_ctl;
 	int threshold = dev_priv->fbc.threshold;
 
+	/* Display WA #0529: skl, kbl, bxt but not for glk*/
+	if (IS_GEN9(dev_priv) && !IS_GEMINILAKE(dev_priv)) {
+		u32 chicken_misc4 = I915_READ(CHICKEN_MISC_4);
+
+		if (i915_gem_object_get_tiling(params->vma->obj)
+				!= I915_TILING_X) {
+                       I915_WRITE(CHICKEN_MISC_4, chicken_misc4 |
+                               FBC_STRIDE_OVERRIDE |
+                               (params->gen9_wa_cfb_stride & FBC_STRIDE_MASK));
+		} else {
+			I915_WRITE(CHICKEN_MISC_4, chicken_misc4 &
+				~(FBC_STRIDE_OVERRIDE | FBC_STRIDE_MASK));
+		}
+	}
+
 	dpfc_ctl = 0;
 	if (IS_IVYBRIDGE(dev_priv))
 		dpfc_ctl |= IVB_DPFC_CTL_PLANE(params->crtc.plane);
@@ -865,6 +880,7 @@ static void intel_fbc_get_reg_params(struct intel_crtc *crtc,
 	struct drm_i915_private *dev_priv = to_i915(crtc->base.dev);
 	struct intel_fbc *fbc = &dev_priv->fbc;
 	struct intel_fbc_state_cache *cache = &fbc->state_cache;
+	int threshold = dev_priv->fbc.threshold;
 
 	/* Since all our fields are integer types, use memset here so the
 	 * comparison function can rely on memcmp because the padding will be
@@ -880,6 +896,17 @@ static void intel_fbc_get_reg_params(struct intel_crtc *crtc,
 	params->fb.format = cache->fb.format;
 	params->fb.stride = cache->fb.stride;
 
+	/* Display WA #0529: skl, kbl, bxt but not for glk*/
+	if (IS_GEN9(dev_priv) && !IS_GEMINILAKE(dev_priv)) {
+		/* calculate cfb stride for y-tiled mode */
+		if (i915_gem_object_get_tiling(params->vma->obj)
+				!= I915_TILING_X) {
+			int cfb_stride = DIV_ROUND_UP(cache->plane.src_w,
+					(32 * threshold)) * 8;
+			params->gen9_wa_cfb_stride = cfb_stride;
+		}
+	}
+
 	params->cfb_size = intel_fbc_calculate_cfb_size(dev_priv, cache);
 }
 
-- 
1.9.1